CFD Turbulence Models Assessment for the Cavitation Phenomenon in a Rectangular Profile Venturi Tube

Abstract

This study investigates cavitation in a rectangular-profile Venturi tube using numerical simulations and four turbulence models. The unsteady Reynolds-averaged Navier–Stokes technique is employed to simulate vapor cloud formation and compared against experimental data. κ-ε realizable, κ-ε RNG, κ-ω SST, and κ-ω GEKO models are evaluated. The simulation results are analyzed for pressure, turbulence, and vapor cloud formation. Discrepancies in cavitation cloud formation among turbulence models are attributed to turbulence and vapor cloud interactions. RNG and SST models exhibit closer alignment with the experimental data, with RNG showing a superior performance. Key findings include significant vapor cloud shape differences across turbulence models. The RNG model best predicts velocity at the throat exit with an error of 4.145%. Static pressure predictions include an error of 4.47%. The vapor cloud length predictions show variation among models, with the RNG model having a 0.386% error for the minimum length and 4.9845% for the maximum length. The SST model exhibits 4.907% and 13.33% errors for minimum and maximum lengths, respectively. Analysis of the cavitation number reveals agreement with the experimental data and sensitivity to cavitation onset. Different turbulence models yield diverse cloud shapes and detachment points. Weber number contours illustrate the variation in the cavitation cloud behavior under different turbulence models.